Writers of Pro Football Prospectus 2008

Most Recent FO Features

LongKyl13.jpg

» Word of Muth: Bears Upset 49ers

The Bears got the win and Kyle Long had the block of his dreams, but Ben Muth found plenty of room for improvement in Chicago's offensive line.

29 Jun 2011

Formation Analysis: Empty Backfields

by Aaron Schatz

Time to take a look at another strategy from the 2010 season: the empty backfield. How often did each team run an empty backfield, and how successful was each team?

Total use of empty backfields was up in 2010, from 3.8 percent of plays in 2009 to 5.0 percent of plays in 2010. The team that used the most empty backfields was quite a surprise: Buffalo. Yes, the team whose strongest position going into the season was running back used nobody in the backfield more than any other team. They weren't particularly successful either. In general, empty backfield plays are very successful, with an overall 16.6% DVOA. Buffalo, however, had just -3.7% DVOA. The Bills were also a bit of an aberration because they were the only team that used empty backfields more than eight percent of the time and didn't make the playoffs.

Empty backfield, of course, doesn't necessarily mean no running backs in the personnel. The Patriots, for example, had Kevin Faulk or Danny Woodhead on the field for nearly all of their empty-backfield plays. The Bills usually had C.J. Spiller or Fred Jackson split out wide, and sometimes both.

I've ranked teams in Yards per Play and DVOA as long as they used empty backfields on at least 2.5 percent of plays.

OFFENSE Pct of Plays Yd/Play Rk DVOA Rk
BUF 15.1% 6.3 7 -3.7% 19
GB 11.0% 5.5 11 29.6% 9
NE 9.9% 7.2 4 78.3% 1
NO 9.6% 8.6 2 77.0% 2
PHI 8.2% 5.5 12 14.1% 13
PIT 8.0% 7.2 5 61.6% 5
ATL 8.0% 5.3 14 35.3% 8
ARI 7.7% 5.9 10 -2.8% 18
SF 6.6% 4.8 18 -32.5% 22
CHI 6.4% 8.2 3 74.9% 3
DET 6.3% 4.8 17 -37.2% 23
HOU 5.7% 4.0 24 7.4% 15
MIA 5.3% 4.9 16 0.9% 17
NFL AVG 5.0% 5.8 -- 16.6% --
WAS 4.9% 3.2 26 -59.7% 25
SEA 4.6% 6.0 8 56.2% 6
DAL 4.4% 7.1 6 72.8% 4
CIN 3.6% 4.4 21 -31.8% 21
STL 3.3% 4.6 19 1.5% 16
KC 3.2% 3.6 25 -44.4% 24
IND 3.1% 5.4 13 29.6% 10
SD 3.1% 4.6 20 -8.0% 20
NYG 2.9% 8.6 1 16.1% 12
CAR 2.6% 6.0 9 7.8% 14
DEN 2.6% 4.2 23 17.1% 11
MIN 2.5% 5.1 15 -68.0% 26
BAL 2.5% 4.2 22 41.6% 7
CLE 2.0% 6.8 -- 9.9% --
TB 1.8% 5.5 -- -34.7% --
NYJ 1.7% 1.8 -- -99.4% --
OAK 1.2% 4.1 -- -24.4% --
TEN 0.7% 2.3 -- -110.7% --
JAC 0.4% 6.0 -- 56.3% --

Posted by: Aaron Schatz on 29 Jun 2011

26 comments, Last at 04 Nov 2011, 12:03pm by Stephen Heller

Comments

1
by Stewart (not verified) :: Wed, 06/29/2011 - 11:00am

Well I never thought I'd see Chicago third in an offensive DVOA ranking.

23
by An Onimous (not verified) :: Tue, 07/05/2011 - 5:40am

Really? You're surprised that Chicago ranks highly in an offensive category when that category absolutely guarantees that Forte cannot possibly run the ball? If anything, this is about the only category I would EVER expect Chicago to rank well in.

Captcha note: my first captcha had a pilcrow (paragraph symbol) buried in the middle. I requested another one, and my new captcha was "tent occupied". Captcha is quite possibly the most entertaining addition to the FO comments since Raiderjoe.

24
by spanky (not verified) :: Thu, 07/07/2011 - 5:14pm

you mean the forte that dropped 4.5 a carry behind a memorably awful o line?

2
by AnonymousA (not verified) :: Wed, 06/29/2011 - 11:12am

Quarterbacks whose teams had over 40% DVOA on empty backfield plays:

Tom Brady
Drew Brees
Ben Roethlisberger
Jay Cutler
Matt Hasselbeck
Tony Romo/Jon Kitna
Joe Flacco
David Gerrard

Gimme a Manning or two for Kitna and Gerrard and you have a pretty good list of "the best QBs in the NFL."

I wonder what the correlation between a QB's DVOA in empty backfield plays and his YPA is?

13
by Intropy :: Wed, 06/29/2011 - 4:51pm

I'd put Rivers above everyone on the list except Brees, and I'd put Rodgers above everyone on the list except Brees and maybe Roethlisberger. You're right that the best empty backfield QBs are also, as you would expect, among the better overall QBs, but I think there's still some data here.

25
by Neoplatonist Bolthead (not verified) :: Mon, 07/18/2011 - 10:05am

What this really means is, "some good passing offenses get advantage from empty backfields and some don't." I don't think there's anything mystical there: Philip Rivers plays in an offense that makes heavy use of runs up the middle, and has a TE who eats up both a LB and a Safety, and still produces. Why would you expect him to perform as well in empty sets as in 3WR sets?

18
by JimmyOz (not verified) :: Thu, 06/30/2011 - 12:02am

Where's the love for the Charlie Whitehurst Experience?

Also, this play in week 4 should be checked to see if it was from an empty set.
3-1-SEA 39 (2nd Q - 2:00) 26-M.Robinson pass deep right to 33-L.Washington to SL 33 for 28 yards (43-C.Dahl)

Can anyone explain why we got DVOA here and not in the 6OL breakdown or explain why this wasn't broken down by QB? I think the reason for leaving this data out is the most interesting thing about the analysis as i can see the formations at the snap.

3
by Tracy :: Wed, 06/29/2011 - 11:18am

Buffalo, Arizona, and NYG, and Minnesota each ranked much better in yards per play than in DVOA out of this formation. Three of those teams were marked by instability at QB. I wonder if this represents a high number of turnovers on these plays, or perhaps these teams had a lot of 8 yard gains on 3rd and 10?

Houston, Denver, and Baltimore, on the other hand, ranked much better in DVOA than in yards per play. My guess is that these teams employed this formation to gain a few yards in high leverage situations. Denver, especially, was probably effected by the Tebow package used in 3rd and 1 or goal-to-go situations throughout the year.

5
by trill :: Wed, 06/29/2011 - 12:27pm

This narrative makes sense to me. I've always thought of empty as an effective choice for teams without a power run game in high-leverage situations, and I think Josh McDaniels agrees with me, but a lot of other coaches use it differently. If you're 3rd and forever, empty gets five receivers into routes very quickly but it telegraphs your protection scheme which is a bad idea when your QB has to wait for someone to run a 15yd dig. And as we established a while back, if you wanna convert 3rd downs with the pass, you gotta throw the ball past the sticks.

I noticed GB and NO going empty quite a bit in the red zone, which would also jive with your analysis (high-leverage, make the defense cover the entire width of the field).

7
by JasonK :: Wed, 06/29/2011 - 1:55pm

Makes sense. The Giants were the "stable QB" outlier in the first group, and they had an unusually high number of turnovers in 2010; it wouldn't surprise me if a number of those happened on Empty-Backfield plays. (Eli's accuracy has always been relatively weak in quick-throw situations, and he had an unfortunate streak of only-slightly-off passes being tipped by the reciever into that hands of a defender early last year.)

16
by PerlStalker :: Wed, 06/29/2011 - 8:28pm

IIRC, Denver usually ran Tebow behind a blocker (badly). He wasn't back there alone on many of his runs early in the season. He was only in for a couple of plays per game, if that, before he took over the starting duties after Orton got hurt.

20
by Mr Shush :: Thu, 06/30/2011 - 2:36pm

Note, though, that the Texans' empty backfield DVOA, while above average, was substantially lower than their overall offensive DVOA, and that they had an outstanding short yardage running game. I'd say the numbers suggest the Texans should be using the empty set less frequently than they do in those situations, and giving Arian Foster the ball more often.

4
by MJK :: Wed, 06/29/2011 - 11:45am

Wow, it's almost like Buffalo didn't have very good coaching last year...

6
by Timmah! (not verified) :: Wed, 06/29/2011 - 12:36pm

I think it might have more to do with having nobody to pick up the defender coming through the turnstile at RT.

9
by trill :: Wed, 06/29/2011 - 3:15pm

With an overall offensive DVOA of -9% and a pass offense DVOA of -5%, one could argue that going empty was the most successful unsuccessful thing that BUF did last year on offense. He's a coach, not a magician.

17
by prophetik (not verified) :: Wed, 06/29/2011 - 11:08pm

i can agree here. gailey did a lot with an average QB and a terrible offensive line last year. most of our starters are another team's scrubs, and they still did some stuff that we haven't seen in years. like pass for 300 yards.

8
by trill :: Wed, 06/29/2011 - 2:18pm

YAY. Empty is my favorite formation, but I would not assume that it's something every team is capable of executing. The defense knows you've only got five (six max, if there's an in-line TE) in protection, so the ball has to come out quickly. If your line can't block theirs 5-on-4, or if your QB holds the ball too long, it won't be pretty.

The best thing about empty IMO is it severely limits what the defense can do. Most D's have one or two checks for empty, and they're usually vanilla coverages with some DL games built-in to generate pressure. If the defense's pass rush is predicated on misdirection/confusion, as some of the best in the league are, empty makes their job a bit tougher, and your OL has a better idea pre-snap of who they're blocking.

IIRC the Bears used empty quite a few times against the Cowboys with great success. Their line was getting torn up on those deeper drops, so they emptied the backfield and got the ball out on a 1-step to Olson on a slant for a TD.

I'd love to see which defenses were the best/worst vs empty. My guess is NYG, MIN, and OAK are in the top five.

10
by Rhombus (not verified) :: Wed, 06/29/2011 - 3:27pm

Your comments on Buffalo are a bit misleading. Buffalo ranked 26th in total offensive DVOA with -10.4%, and 26th in passing DVOA with -5.0%. This suggests that Buffalo's empty backfield formations were generally MORE successful than their typical offensive production.

12
by trill :: Wed, 06/29/2011 - 4:35pm

That's true, but the league-average DVOA out of empty is 16%, so league-wide teams were 16% more successful out of empty than they were on the average play from scrimmage. If BUF's empty DVOA was, say, +11% or even +5%, then we could say that they were at least on pace with the rest of the league regarding that particular formation. As it stands they were better than their usual selves, but still behind the curve.

21
by Mr Shush :: Thu, 06/30/2011 - 2:37pm

Not so: 2010 league average offensive DVOA was well above 0%.

11
by DisplacedPackerFan :: Wed, 06/29/2011 - 4:29pm

So I've read a few places that Green Bay used the 5 wide out empty backfield set more than anyone else (something like 31 times). I also know they did stuff with motioning players out of the backfield. I'm curious if you have the numbers on how they did in the line up empty vs the motion out to empty. I know sample size is going to start getting pretty small with that, but I'm curious as to what DVOA might have been for those sets.

14
by Aaron Schatz :: Wed, 06/29/2011 - 5:21pm

No, sorry. We only record formation and personnel as of the snap. Trying to figure out how to mark motion would add another whole element that would take the volunteers extra time.

22
by Kevin from Philly :: Fri, 07/01/2011 - 2:23am

Hey, no getting soft on the lackeys!

15
by Eric G (not verified) :: Wed, 06/29/2011 - 5:44pm

Was the success of the empty backfield due to the shotgun? What was the DVOA of teams that used the shotgun and had at least one running back?

19
by bubqr :: Thu, 06/30/2011 - 3:18am

EVIL. EVIL. EVIL.

26
by Stephen Heller (not verified) :: Fri, 11/04/2011 - 12:03pm

I want to ask you about the performance of Cardenals in this last season, do you really believe they deserved win this year? and what about that rumor, about they use a method know as 4rx? please if you can answer me I gonna appreciate it.